Jump to content

Cable Run from Profitable FX


Recommended Posts

Hi teknomage the thing is still doing error 130. I'm on live in a little bit of DD. Tried to P.M you but can't get through. The demo running along side is fine. Is there something I should know from the broker. Trades have opened and not opened. Big gap between price and last trade. Broker will get a call today. Maybe their fault maybe not. On 0.01 lots but still very concerned. Should make the effort and backtest. Another forex lesson.
Link to comment
Share on other sites

  • Replies 921
  • Created
  • Last Reply

Top Posters In This Topic

Are you sure it's FXOpen ECN data?

FXOpen ECN is simply Dukascopy with MetaTrader.

If you had downloaded the data from the FXOpen terminal History Center then the data is probably from MetaQuotes, which is very poor quality data, and not from Dukascopy.

 

See my thread here: http://indo-investasi.com/showthread.php/9257-Holes-in-MetaQuotes-historical-data-(via-Alpari-UK-History-Center)-DO-NOT-USE

 

Hi teknomage, great efforts here thanks.

 

While backtesting on Metaquotes preset 3 for 2010 I do get quite similar results as yours. Looking into the details there is a 6 month gap as below. This problem was mentioned prior already in this thread from others on the original cable run base and I put it aside that time. Will try a run on tick data see how that goes...

 

http://img232.imageshack.us/img232/8526/31723157.jpg

 

Cheers

Link to comment
Share on other sites

I think the iCustom calls are unnecessary, as it simply gets ATR values divided by the point value. iCustom requires more processing and so would make backtesting slower.

 

So you could replace the lines

  double ATRPips =iCustom(NULL,PERIOD_M15,"ATR_Pips",ATR_Period,0,0);
  double ATRPrePips1 =iCustom(NULL,PERIOD_M15,"ATR_Pips",ATR_PrePeriod,0,ATRShift1);
  double ATRPrePips2 =iCustom(NULL,PERIOD_M15,"ATR_Pips",ATR_PrePeriod,0,ATRShift2);

 

with the lines:

  double ATRPips =iATR(Symbol(),PERIOD_M15,ATR_Period,0) / point;
  double ATRPrePips1 =iATR(Symbol(),PERIOD_M15,ATR_PrePeriod,ATRShift1) / point;
  double ATRPrePips2 =iATR(Symbol(),PERIOD_M15,ATR_PrePeriod,ATRShift2) / point;

 

You should also put this code to determine appropriate point value for 5 digit brokers in init():

  double point=Point;
  if (point==0.00001) point=0.0001;
  else if (point==0.001) point=0.01;

 

There would then be no need for the ATR_Pips indicator in backtesting.

Link to comment
Share on other sites

I think the iCustom calls are unnecessary, as it simply gets ATR values divided by the point value. iCustom requires more processing and so would make backtesting slower.

 

So you could replace the lines

  double ATRPips =iCustom(NULL,PERIOD_M15,"ATR_Pips",ATR_Period,0,0);
  double ATRPrePips1 =iCustom(NULL,PERIOD_M15,"ATR_Pips",ATR_PrePeriod,0,ATRShift1);
  double ATRPrePips2 =iCustom(NULL,PERIOD_M15,"ATR_Pips",ATR_PrePeriod,0,ATRShift2);

 

with the lines:

  double ATRPips =iATR(Symbol(),PERIOD_M15,ATR_Period,0) / point;
  double ATRPrePips1 =iATR(Symbol(),PERIOD_M15,ATR_PrePeriod,ATRShift1) / point;
  double ATRPrePips2 =iATR(Symbol(),PERIOD_M15,ATR_PrePeriod,ATRShift2) / point;

 

You should also put this code to determine appropriate point value for 5 digit brokers in init():

  double point=Point;
  if (point==0.00001) point=0.0001;
  else if (point==0.001) point=0.01;

 

There would then be no need for the ATR_Pips indicator in backtesting.

 

Very good solution, since the ATR Data is now available from the start for all varibales (ATRPips and PrePips). Noticed, that the ATR Pips Filter needs to "feed" the Array first on the defined bars to get the PrePips data availbale.

 

Cheers

Link to comment
Share on other sites

Hi teknomage and everyone,

I've been getting error 130 (even on latest mod by teknomage). I spoke to my broker Go markets and asked about stop levels and a few other things upon researching error 130 on the internet. The broker told me that my problem is that my Ea is set to use instant execution for orders where they are market execution. Alpari is instant execution. Is there a way to change the coding to accept market execution orders or should I just change brokers? Any one please as this is important cause I've started trading live and I don't want my account to go Kaput!

I'm going to browse the entire thread again and hopefully I might pick up some useful info... @-)

 

My error 130 says this:#988675 Error: 130 Function: New Order (3,0.01,1,5851,0,0)36

 

Is it still because of invalid stops?. If anyone knows a bit more please don't hesitate to respond as this is also helps the entire thread. My range is set at 17. Do I change this to 170?

 

I also checked my broker stop level and it is 0 pips so the stop level is close but apparently according to teknomage it's not a problem.

 

Has this coding been utilised? Post #446. I had a look inside the coding. I wouldn't know where to put it.

 

If you still got error 130, try this:

gd_124 = FindLastSellPrice(Magic);// GetLastSellPrice(Magic);

gd_132 = FindLastBuyPrice(Magic);// GetLastBuyPrice(Magic);

 

if (gd_124 < Bid) gd_124 = Bid;

if (gd_132 > Ask) gd_132 = Ask;

 

Thanks for responding everyone. I've just skimmed through the entire thread.Some options. My broker is market execution so will things change if I go to instant execution broker? Can some coding help here? Or do I just let go of CableLarikan ok 24H and move on to the other cool Ea's on this thread? Teknomage the mod version prevent errors is still doing so but it just might be an inherent problem with CableLarikan.

Edited by tomislav
Link to comment
Share on other sites

Hi teknomage and everyone,

...

Alpari is instant execution. Is there a way to change the coding to accept market execution orders or should I just change brokers? Any one please as this is important cause I've started trading live and I don't want my account to go Kaput!

...

 

Have a look at this document please. I'm not a coder but it gives some code on page 6. Although it doesn't talk about buy/sell limits, it might be helpful for you.

 

http://www.bostontechnologies.com/docs/FAQ.pdf

 

John

Edited by TheBeast
Link to comment
Share on other sites

I demoed CableLarikan prevent errors on another broker a reputed ECN with instant market execution and got error 4105 or whatever it is. It's just a flawed EA for 5 digit brokers and because it can be very dangerous to say the least. Disappointing because of the immense potential and the only barrier is a few lines of code that need to be implemented. Edited by tomislav
Link to comment
Share on other sites

I've just finished a visual backtest from 2008-07-31 and it wiped out a 100K USD account! There were 79 trades open until account stop out.

Here are the environment details:

EA name: CableMuruku24H_teknomage

set file: cr24h_teknomage - optimized03

Symbol: GBPUSD

Time frame: H1

Start date: 2008-07-31

End date: n/a

Initial deposit: 100000 USD

Leverage: 500

Stop out level: 20%

 

Here is a screen capture of the visual backtest chart a short while before the account stopped out (the green horizontal dashed lines represent open trades (plus a couple of pending limit orders)):

http://stashbox.org/1015593/gbpusd_h1_cablemuruku24h_teknomage_set%3Dcr24h_teknomage%20-%20optimized03_2008-07-31.gif

 

 

It seems like the forum shrinks images. Visit this URL for the original image: http://stashbox.org/1015593/gbpusd_h1_cablemuruku24h_teknomage_set%3Dcr24h_teknomage%20-%20optimized03_2008-07-31.gif

Edited by hyperdimension
Link to comment
Share on other sites

Well guys, my first two nights using set file #3 went great, but then the bull trend hit cable last knight and took a $11K demo account down to $2000 overnight. I was just testing the system using 1 full lot which I know is too high, but just wanted to see what would happen. This EA seems to work great in sideways markets, but strong trend seem to be killers.
Link to comment
Share on other sites

I've just finished a visual backtest from 2008-07-31 and it wiped out a 100K USD account! There were 79 trades open until account stop out.

Here are the environment details:

EA name: CableMuruku24H_teknomage

set file: cr24h_teknomage - optimized03

Symbol: GBPUSD

Time frame: H1

Start date: 2008-07-31

End date: n/a

Initial deposit: 100000 USD

Leverage: 500

Stop out level: 20%

 

Here is a screen capture of the visual backtest chart a short while before the account stopped out (the green horizontal dashed lines represent open trades (plus a couple of pending limit orders)):

http://stashbox.org/1015593/gbpusd_h1_cablemuruku24h_teknomage_set%3Dcr24h_teknomage%20-%20optimized03_2008-07-31.gif

 

 

It seems like the forum shrinks images. Visit this URL for the original image: http://stashbox.org/1015593/gbpusd_h1_cablemuruku24h_teknomage_set%3Dcr24h_teknomage%20-%20optimized03_2008-07-31.gif

 

as with all martingales or grids, they can get on the wrong side. so one has to manually intervene at times and take whatever loss and hopefully make it up at a later time. its all about money management, running a small lot size is always the way to go. dont let greed rule, or account kaputt.

Link to comment
Share on other sites

...

It's just a flawed EA for 5 digit brokers and because it can be very dangerous to say the least. Disappointing because of the immense potential and the only barrier is a few lines of code that need to be implemented.

Tomislav, sadly I have no clue how to code but I have found something that might help you:

 

http://forum.mql4.com/30672

 

John

Link to comment
Share on other sites

@ peterke and TheBeast : Thanks for sharing info, but there is no TP or SL in any of these CR mods, except for maybe Expat's version. So, there is no need for a separate OrderModify to specify the TP or SL...

 

@ hyperdimension: Thanks for all your input and sharing your results. What do you suggest to improve this robot's performance, besides what I've mentioned below?

 

@ newbee & reinerh: Yes, you're both right. So, the best thing we can hope to do is filter out strong trends. I will be working on adding a News filter to my next version. In the meantime, if any of y'all can suggest good indicators or indi-combos that can help us to prevent the robot from trading during impending runaway trends, that'll be great. I'll try to incorporate them into the next version.

 

The SL and TP need to be sent in separate transactions after the buy or sell order.

rgds

Peter

Link to comment
Share on other sites

Last night was a classic example of going against a very bullish trend. CableRun took a sell trade just as the 7 day MA was just about to cross the 200 day MA, with the 13 day MA not far behind it. Within a couple hours, both 7 and 13 MAs then also crossed the 100 day. By the time the 13 day crossed the 200, it was all over, but CB just kept selling. Possibly some sort of MA cross filter could be incorporated. I also wonder if the Black Dog system could be incorporated, were CB was restricted from taking any counter trend buy or sell that goes against the Black Dog primary signal. Just a few thoughts.
Link to comment
Share on other sites

I'v running Cable Run for three weeks - started with 1000 deposit and now Balance is at 8776 USD and eq 5565 USD. I'm running two versions with 0.1 lots.

I'm using manal closing and hedging - last run closed my hedge with profit now I do not need to hedge before next run - maybe before Nov 3th.

 

Hope I'll be able to trade this way at least 2-3 months before I'l go live. Anyway last two years are showing that EAs are failing in Oct 2q to November end. Well see how it goes.

Link to comment
Share on other sites

Isn't that called being a martingale? I've chucked Mr. CableLarikan into the dust bin. Yesterdays up trend happened during GMT afternoon/evening/next morning. Damage done X_X There was a GBP news event at 8:30 GMT (claimant count change) and a US news event at 8:10 GMT (Fed Speech). If one keeps to watching upcoming news events then no trades should have been placed for a day. How was one to know that the impact of these two high impact news events would carry a relentless upswing for 18 hours. When the GBP moves it REALLY moves. News filter would be great. I'm demoing the teknomage versions and for me the original CableLarikanMuruku v5 teknomage is working beautifully.

No bugs!

 

The recent teknomage has inconsistencies when placing trade baskets and has gaps in the trade baskets. This is what happens to me. I'm getting to believe that these Cable Run EA's are way too broker dependant or there are still some bugs. They should work the same. Can you add a news filter to the CableLarikanMuruku v5 teknomage just for me or anyone else who finds it useful? V5 teknomage doesn't work the same as the later mod for me. There are inconsistencies. I'm guessing that news filter is FFCl from Forex Factory? It's a good one.

 

One way to avoid break away trends is to not trade when there are high impact GBP and US news events just 20 min apart. And pay particular attention to high impact GBP news especially. GBP is renowned for massive movements that can destroy or fill up accounts. Set your news filter to not trade for 18hours or so after the fact and 4-6 hours before. Set the news filter to avoid high impact GBP news events/speeches only. The US news appears to not launch GBP/USD as much as GBP news.

News filter on V5 (as well) please and I think they would be trade ready products. >:D<

 

http://www.4shared.com/file/ICipydcW/FFCal.html

 

This a news filter I got from XUX99; a programmer at the Forex factory. Works perfectly.

 

P.S I'm demoing latest teknomage mod with AXI trader (ECN insant execution) and even though there is an occasional invalid SL TP message in the journal it is working perfectly on this broker at this point in my testing. I can't change brokers just yet so thanks for any help teknomage. It's a big job but it's almost there.

Edited by tomislav
Link to comment
Share on other sites

@ peterke and TheBeast : Thanks for sharing info, but there is no TP or SL in any of these CR mods, except for maybe Expat's version. So, there is no need for a separate OrderModify to specify the TP or SL...

 

@ hyperdimension: Thanks for all your input and sharing your results. What do you suggest to improve this robot's performance, besides what I've mentioned below?

 

@ newbee & reinerh: Yes, you're both right. So, the best thing we can hope to do is filter out strong trends. I will be working on adding a News filter to my next version. In the meantime, if any of y'all can suggest good indicators or indi-combos that can help us to prevent the robot from trading during impending runaway trends, that'll be great. I'll try to incorporate them into the next version.

 

Certainly, we need to look further to improve the safety of this strategy. The strategy itself appears pretty good to me as it is stable over 10 years. What appears important is when the strategy should be trading (Filters), how it manages the trades and how can it recover while in trouble however.

 

We dont need to look too far, Muruku posted prior his latest filter settings. That helped a lot and some bugs fixed in my version shared prior... 10 year backtest was going thru, well it took 4 days...

 

Newsfilter is an option (you just might take FFCal from the Megadroid N1... copy/paste...) where I am looking into to get the newsdata downloaded to make it backtestable... cant recall who did or mentioned this prior...as well as a permanent at least secondary Filter (e.g. EMA improves performance but I guess that can be driven further on others or combined). But lets not filter it to dead ;)

 

Besides Filters, we need to look in the trading strategy. E.g. Filters might not work if we want or need to stay on trading while a basket is open.

 

Try with different/less TotalProfit settings. You might see the dd melting... Less dd/open trade time in this EA keeps it more open for trading thus the overall Profit might balance out or even improve.

 

What I have set as option is e.g. to reduce the Total Profit on each next trade from a certain number of open trades onwards. That drives further for less dd and as well the overall Profit might just balance out while the strategy is ready for trading again much quicker.

 

Next trading strategies on my list is to try something like a multiple basket option and a combimation of Cable Run scalping/adding breakout and/or trending (like Forex Combo which applies its 3 strategies nicely while in trouble to balance out as much as possible...) see how that goes.. That could be the parachute for this strategy.

 

I have learned a lot of MQL in the last two weeks. My last programming I did 20 years ago on Pascal... Anyway, it will help me a lot in my further trading with MT4 and I just can encourage you guys to give it a try. Get started with EA Generator, build a logic then you can look into the code try to modify to your further needs...

 

Cheers

Link to comment
Share on other sites

tomislav,

 

i think the big move was due to the usa selling a bunch more bonds last night asian time. it made the usd plunge.

and it will keep plunging for a while longer i am sure, that means cable euro and gold will keep going up.

our greenback will be worthless before too long, sigh.

Link to comment
Share on other sites

You're right. GBP rose against the AUD (hint) but no where near to the unprecedented rise against the Greenback. News filter will help for sure however planning and being informed are crucial to trade the GBPUSD.

Just checking my demos... yeah teknomage v5 is rock solid. Thanks again teknomage.

P.S There you go kerozin.. all the little GBP piggies are coming home.

Edited by tomislav
Link to comment
Share on other sites

Im running GU, UJ and UCHF on a 10k account with 0.1 lots. After 4 days have $12,254 in balance and $1850 in unclosed trades. I am using V5 technomage version. I had a maximum $6500 draw down yesterday and have been manually closing some trades.

I have been running 24hrs after the first day.

Running opposing pairs does not really work as the pairs seem to double the risk rather than reduce the risk.

I am thinking the best way to deal with a trend will be to manually hedge to limit the drawdown.

 

Maybe some logic along the lines of

when drawdown = x% of balance (or a fixed $amt) enter an opposing trade for the net pips

either have a manual exit of hedge or auto close

 

By setting the max drawdown to say 20% of account balance the account is at least preserved

 

rgds

PEter

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.




×
×
  • Create New...